ON ELECTROLYSIS. 355 



2. Again, it appears to me that the atomic dissociation hypothesis requires that 

 the majority of compounds, if not all, should ^;e;- se conduct more or less well, 

 especially if it be admitted that ionic velocities differ ; as a matter of fact, however 

 only a very limited number can be regarded as electrolytes. Moreover, according 

 to the orthodox view — particularly in the form in which it is stated by Arrheuius 

 — the most active substances are those which contain the greatest number of active 

 atoms, i.e., those which are most easily dissociated. But, in point of fact, the sub- 

 stances which are most active — both chemically and electrically — are by no means 

 those which we should regard as most likely to dissociate ; thus, HCl, HBr and 

 HI differ in stability to a very marked extent, yet their molecular conductivities 

 in aqueous solution are almost identical. And a glance through the list of salts 

 which are believed to be per se electrolytes is sufficient to show that these are not, 

 as might fairly be expected, among the leastst^able, but quite the contrary : silver 

 chloride, bromide and iodide, for example — all compounds of considerable stability 

 — being the best conductors known, I believe, among simple electrolytes. 



3. Aqueous alcoholic solutions generally oppose a greater resistance than the 

 corresponding aqueous solutions, and solutions in absolute alcohol oppose a 

 practically infinite resistance ; yet sm-ely it might be expected that alcohol — indeed, 

 any neutral solvent— woidd screen the dissociated atoms from each other and thus 

 render electrolysis possible. 



4. If, as Professor Lodge asserts, the orthodox view (or his version of it) 

 pictures the dissociation of water by salt quite as easily as that of salt by water, 

 why is conduction assumed by Kolilrausch and others to take place only through 

 the agency of the atoms of the dissociated salt, and not at all through that of the 

 water, except, perhaps, to judge from a recent admission made by Kohlrausch, in 

 the case of verij dilute solutions ? Certainly, on the atomic dissociation hypothesis, 

 both water and salt, I imagine, are to be regarded as dissociated ; and, moreover, it 

 would appear probable that as the dissociated constituent atoms of water 

 would have less chance in a concentrated solution of coming together aoain con- 

 duction would take place mainly through their agency; and that in a dilute'solu- 

 tion, for a similar reason, conduction would be affected chiefly throuo-h the 

 agency of the salt. This conclusion is manifestly opposite to that arrived" at by 

 Kohlrausch. 



5. Arrhenius asserts that the conductivity of ammonia solutions is caused by a 

 small qiiantity of NHpH, which is increased by dilution ; and in reply to Pro- 

 fessor Lodge's remark deprecating this statement (B.A. 'Report,' 1886, p. 363), he 

 says : ' I may say in explanation that almost all chemists attribute the reaction of 

 ammonia to a small portion of NH^OH in it, &c.' I am one of those chemists who 

 think that it is not necessary to make this assumption, believing that the chemical 

 changes produced by ammonia solution are due to the combined or simultaneous 

 action of ammonia and water, but I quite agree with Arrhenius that the effect of 

 dilution is to increase the proi^ortion of simple or active molecules i.e. to dis- 

 sociate^ the molecidar complexes into the constituent simple molecules of ammonia. 

 The opinion is common among chemists that the nitrogen becomes separated froni 

 the hydrogen not by direct electrolysis, but by means of oxygen primarily pro- 

 duced by electrolysis — in other words, that water is the electrol-vi;e. 



6. The dissociation hypothesis has i:ot only found favour with physicists, but 

 also with chemists, as it long seemed to afford a simple explanation of the occur- 

 rence of chemical change ; in fact, the popular view may be summed up in the 

 simple statement that simplification as a rule precedes complication. As I have 

 more than once insisted, however, recent observations on the influence of minute 

 amounts of third substances oblige us to admit that we have yet much to learn 

 regarding the manner in which apparently the simplest changes occur. In certain 

 cases where we may almost assert that we know dissociated atoms to be present, 

 combination or mterchange does not ensue unless a minute amount of a third 

 apparently neutral substance be present : indeed, the great problem in chemistry 

 which is but now being attacked, is whether it is possible for chemical chano'e to 

 occur between any tivo substances, be they simple atoms or more or less complex 

 molecules. Even dissociation would seem not to be a simple function of tempera- 



A A 2 



